Focusing on the vertical array source signal recovery method, aiming at the estimation deviation caused by environment mismatch in the engineering application of virtual time reversal mirror technology, this paper proposes the normalized virtual time reversal technology through theoretical derivation and analyzes the performance of the virtual time reversal technology under the change of environment and signal-to-noise ratio through simulation. Finally, the practicability of the normalized virtual time reversal method and the effectiveness of model correction are verified through experimental data. The data processing results show that the source level estimation error of vertical array based on normalized virtual time reversal algorithm is less than 5.2dB. In the experimental environment of this paper, the correlation of signal spectral structure recovered by this method is about 5% higher than that by the traditional method.

Aiming at the extraction of failure character signal for early acoustic emission (AE) signals, a method to combine lifting wavelet packet with spectrum zoom technology is developed. Using lifting wavelet packet de-noising method, not only filters the noise signal, but also retains the feature information. After the signals de-noising, combining the envelope demodulation analysis with the spectrum zoom technology, which can effective extract the weak fault characteristic frequency The result of simulated signals and experimental signals shows that the proposed method has better noise reduction with a higher signal noise ratio (SNR) and lower mean square error (MSE), and it can successfully extract the fault characteristic frequency of rotating machinery early AE signals.